Applying responsible AI practices in KinderLogger - OpenAI API Tutorial

Contents

• Introduction Introduction

  • Building a real-world AI app

    34s
  • What you need to know

    1m 48s
  • What you are going to build throughout this course

    2m 29s
  • Preparing the tools and development environment

    1m 35s
• 1. Using the Audio API to Transcribe Audio and the Chat Completions API to Create Documents 1. Using the Audio API to Transcribe Audio and the Chat Completions API to Create Documents

  • (Locked)
    Extracting rich information from audio messages

    37s
  • (Locked)
    Installing the required library and preparing the code

    1m 53s
  • (Locked)
    Transcribing the audio files by using the Audio API

    3m 29s
  • (Locked)
    Creating the JSON document using the Chat Completions API

    3m 56s
• 2. Building the KinderLogger Assistant with the Assistants API 2. Building the KinderLogger Assistant with the Assistants API

  • (Locked)
    Bringing the KinderLogger assistant idea to life

    30s
  • (Locked)
    Creating the assistant using the Assitants API

    4m 8s
  • (Locked)
    Creating a new conversation thread and adding a message

    1m 53s
  • (Locked)
    Running and testing the conversation thread

    2m 39s
• 3. Implementing a Retrieval Augmented Generation (RAG) Process in the Assistant 3. Implementing a Retrieval Augmented Generation (RAG) Process in the Assistant

  • (Locked)
    Using private documents and data in the KinderLogger project

    30s
  • (Locked)
    Uploading the transcription files to the assistant

    1m 54s
  • (Locked)
    Improving the assistant with prompt engineering techniques

    2m 49s
  • (Locked)
    Using the assistant on the Assistants Playground

    1m 26s
• 4. Implementing the Web API to Expose the Assistant 4. Implementing the Web API to Expose the Assistant

  • (Locked)
    Exposing the assistant as a service

    35s
  • (Locked)
    Creating the Web API using the FastAPI library

    2m 46s
  • (Locked)
    Calling the assistant from the Web API

    2m 40s
  • (Locked)
    Implementing logging in the Web API

    3m 9s
  • (Locked)
    Building and publishing the Docker container image

    2m 13s
  • (Locked)
    Deploying the Azure Container App for the service

    3m 22s
• 5. Identifying Harmful Content Using the Moderations API 5. Identifying Harmful Content Using the Moderations API

  • (Locked)
    Applying responsible AI practices in KinderLogger

    28s
  • (Locked)
    Implementing the Moderations API in the transcription app

    1m 7s
  • (Locked)
    Implementing the Moderations API in the Web API

    2m 7s
  • (Locked)
    Building a new image and updating the Azure Container App

    1m 48s
• 6. Creating the KinderLogger GPT 6. Creating the KinderLogger GPT

  • (Locked)
    Using a GPT as a frontend for the KinderLogger project

    28s
  • (Locked)
    Creating the GPT and the action for invoking the service

    3m 28s
  • (Locked)
    Testing the KinderLogger GPT

    2m 24s
• Conclusion Conclusion

  • (Locked)
    Next steps

    1m 2s